Time
3 months
Location
USA
Sector
Cybersecurity / Banking
Description
The client required a system for risk analysis tailored to financial organizations. The project’s goal was to develop a system that combined machine learning and rule-based approaches for risk estimation based on numerous factors within the company. Additionally, it aimed to track employees’ actions within the organization and identify intruders based on unusual behaviors or abnormalities.
Solution
A large-scale system for risk analysis and detection was created within the company. Real data from various datasets related to organizational activities, such as the Enron dataset, were utilized to simulate organizational dynamics, including emails between employees and business processes. Initially, a synthetic dataset was used before the first customers came onboard. The solution involved the creation of a rule-based system designed to identify potential intruders and detect data anomalies. Machine learning models were also trained to generate risk scores for each employee in the organization.
Results
The implementation of this proof of concept was highly successful. The client was able to onboard several clients, including large corporations, which subsequently led to the development of a full-scale application. The client expressed great enthusiasm for the proof of concept, which served as the foundation for transforming it into a fully realized product. This initiative marked a significant step in providing financial organizations with a powerful tool for risk analysis and employee behavior tracking.